oracle.dmt.odm.settings.algorithm
Class SVMRegressionSettings

java.lang.Object
  |
  +--oracle.dmt.odm.MiningObject
        |
        +--oracle.dmt.odm.settings.algorithm.MiningAlgorithmSettings
              |
              +--oracle.dmt.odm.settings.algorithm.SupportVectorMachineSettings
                    |
                    +--oracle.dmt.odm.settings.algorithm.SVMRegressionSettings

All Implemented Interfaces:

java.io.Serializable

public class SVMRegressionSettings

extends SupportVectorMachineSettings

An instance of SVMRegressionSettings is used to specify settings for the Support Vector Machine (SVM) mining algorithm to build a regression mining model. Specifies parameters specific to regression using SVM.

Since:

10.1.0

See Also:

Serialized Form

Constructor Summary

SVMRegressionSettings()
Creates an instance of SVMRegressionSettings object with default settings: - default normalization is set to minMax - default kernelFunction is set to linear - default tolerance value is set to 0.001 - default complexityFactor value is set to 1 - default kernelCacheSize value is set to 50,000,000 - default epsilon value is set to 0.1

SVMRegressionSettings(Normalization normalization, KernelFunction kernelFunction, float tolerance, java.lang.Float standardDeviation, java.lang.Float complexityFactor, int kernelCacheSize, java.lang.Float epsilon)
Creates an instance of SVMRegressionSettings object.

Method Summary

java.lang.Float	getEpsilon() Returns the interval of the allowed error in epsilon-insensitive regression.
void	setEpsilon(java.lang.Float epsilon) Sets the interval of the allowed error in epsilon-insensitive regression.

Methods inherited from class oracle.dmt.odm.settings.algorithm.SupportVectorMachineSettings

getComplexityFactor, getKernelCacheSize, getKernelFunction, getNormalization, getStandardDeviation, getTolerance, setComplexityFactor, setKernelCacheSize, setKernelFunction, setNormalization, setStandardDeviation, setTolerance

Methods inherited from class oracle.dmt.odm.settings.algorithm.MiningAlgorithmSettings

getMiningAlgorithm, getMiningAlgorithmName

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SVMRegressionSettings

public SVMRegressionSettings(Normalization normalization,
KernelFunction kernelFunction,
                             float tolerance,
                             java.lang.Float standardDeviation,
                             java.lang.Float complexityFactor,
                             int kernelCacheSize,
                             java.lang.Float epsilon)
                      throws InvalidArgumentException

Creates an instance of SVMRegressionSettings object.

Parameters:

normalization - value of normalization

kernelFunction - value of kernel function

tolerance - value of tolerance

standardDeviation - the standard deviation of the Gaussian kernel this field is required if kernelFunction is Gaussian. The value will be computed if null, and the computed value can be obtained from a restored model.

complexityFactor - the parameter trading off complexity for prediction loss. The value will be computed if null, and the computed value can be obtained from a restored model.

kernelCacheSize - the parameter of kernel function cache size

epsilon - value of the interval of the allowed error in epsilon-insensitive regression

Throws:

InvalidArgumentException
- - normalization is null
- kernelFunction is null
- tolerance <=0
- standardDeviation <=0 and standardDeviation !=null
- complexityFactor <=0 and complexityFactor !=null
- epsilon <=0

InvalidArgumentException

SVMRegressionSettings

public SVMRegressionSettings()

Creates an instance of SVMRegressionSettings object with default settings: - default normalization is set to minMax - default kernelFunction is set to linear - default tolerance value is set to 0.001 - default complexityFactor value is set to 1 - default kernelCacheSize value is set to 50,000,000 - default epsilon value is set to 0.1

Throws:

InvalidArgumentException
- - kernelFunction is null

Method Detail

setEpsilon

public void setEpsilon(java.lang.Float epsilon)
                throws InvalidArgumentException

Sets the interval of the allowed error in epsilon-insensitive regression.

Parameters:

epsilon -

Throws:

InvalidArgumentException - is thrown
- when epsilon <= 0

InvalidArgumentException

getEpsilon

public java.lang.Float getEpsilon()

Returns the interval of the allowed error in epsilon-insensitive regression.

Returns:

m_epsilon